Scary But True: Real Reform is Wholesale

Thanks go to Forest Hinton at The Quick and the Ed for his Quick Hits today leading me to Heather Zavadsky’s report for the American Enterprise Institute for Public Policy Research, “Embracing Systemic Reform.” In the last few years, there have been many education advocates pushing for individual policies, such as performance pay and breaking up large schools into smaller ones. However, AEI argues that true reform can only happen when a web of changes is made. They point out a few reasons for this.

Spending green isn't useful unless it's strategic.

Most importantly, students are diverse, so not every student will respond to any one change. Smaller classrooms may be useful for students who need more attention, but it doesn’t help the student who is ready to move beyond the curriculum being taught. Individual policies will almost never affect the entire population. Next, policies affect each other. What is the point of giving teachers bonuses for helping students achieve higher test scores when they are stuck to a misaligned curriculum, or worse when the test scores themselves are meaningless because of poorly-written assessments? The idea of creating systemic change is what has inspired so many CMOs to establish their own schools and districts. It’s also the reason that last week, I said that Bill Gates should invest in an entire charter district. The only way to help all of the students in a district is to alter the way schools operate entirely – and all of the schools a child attends K-12 for that matter. Only an entire district has the ability to control all of the aspects of a student’s learning. The report focuses on five large school district who have managed to improve learning outcomes across the board for their students, in some cases significantly reducing the achievement gap between minority and white or low- and high-income students. Brookings has similar findings in the commonalities of these districts to a report I wrote for The Riordan Foundation. Who would have thought that using data, aligning your curriculum, building a skilled staff, and having strong goals would be the keys to strong schools?

On the flipside, just because you’re spending lots of money on bold reform doesn’t mean all of your changes are doing anything. The Brookings Institute has found that the services that the Harlem Children’s Zone has supported have had no educational effect. That doesn’t mean that the schools themselves have not been effective. It’s just that students receiving the additional social services, such as classes for parents, healthcare, and after-school programs, did not show any improvements over those who didn’t. Not all problems are strong factors in education. Perhaps Obama shouldn’t have been so hasty in endorsing $210 million for Promise Neighborhoods to replicate the HCZ.

Creating the Finish Line

It seems as though I am not alone in my criticism of Cortines from my last post. The mayor’s not too fond of his mettle, either. He says he’s not supportive enough of charter schools.

As I said on Thursday, one of the big recent developments is Race to the Test, a contest to create assessments that align with the new Common Core standards. Personally, if they can actually make legitimate assessments that aren’t just multiple-choice fill-in-the-bubble exams that force students to regurgitate information for a few hours at a time, not only will I be impressed, but it will be a major step forward from what the current state tests look like.

The odd thing about this particular race, though, is that instead of allowing a number of different groups to compete for the prize, only three organizations were tapped and all three will be getting something. According to EdWeek, there were originally six consortia, but because there was so much overlap, they combined forces. Two of the three consortia, consisting of 26 and 31 states, are competing for $320 million of the total $350 million to create tests for all grades, while a smaller group of 12 states is aiming at making reliable high school exit exams. Even though these are states creating the assessments, it seems as though the federal government’s suggestions may go a long way to shaping what the tests look like. The SMARTER Balanced Assessment Consortium and the Partnership for the Assessment of Readiness for College and Careers initially started out with major differences in their proposals, but after the receiving comments on them, their plans now look very similar to one another. Both are planning on having performance assessments spread throughout the year to track development along with a big exam at the end.

Perhaps smarter tests can be integrated into schools better.

One major difference is that despite the fact that both groups are using technology, SMARTER seems to have latched onto the computer-adaptive model, which can lead to greater accuracy and faster tests. For those unfamiliar, most graduate entrance exams (the LSAT, the GRE, etc) at this point use computer-adaptive tests that change depending on how well the student is doing. They are essentially the exam equivalent of an optometrist trying to figure out your vision. Instead of changing the strength of the prescription depending on how well you can see,  questions get harder or easier depending on how well you perform on each question. That way, they can zero in on where your performance is. Why is this important? In most states, standardized tests take a significant amount of time. I can remember my own high school experience in Indiana when the entire school would essentially shut down for a week for the sophomores and students who hadn’t previously passed to take the ISTEP. As a teacher in Arizona, there were separate days  for reading, writing, math, and science sections of AIMS, as well as an additional day for freshmen to take the TerraNova, and that doesn’t even include the extra days in the fall for students who failed the previous spring to retake the tests. If these tests used adaptive technology, they could be whittled down to a fraction of the length, resulting in more accuracy because of less burn-out from the sheer length of the tests.

While these developments are encouraging, there has been some criticism of the process. EdWeek’s Catherine Gewertz says that the timeline might be too quick. The DOE wants these tests to be in use by 2014-2015. For a strong test to be administered, there needs to be considerable piloting and adjustment and four years may not be enough to reliably do that. Bill Tucker of The Quick and the Ed and EducationNext warns that there are some important steps that may or may not happen that could determine the success of the process. His biggest concern is having open platforms and shared infrastructure. The two consortia seem to be working together so far, so it looks like a positive start. We won’t know how successful this initiative is for a couple of years probably, though.